Purpose: The aim of the project is to test the hypothesis that quasi‐breath‐hold (QBH) biofeedback improves the residual respiratory motion management within a gating window, reducing respiratory motion artifacts in gated 3D thoracic MR images. Methods: A QBH biofeedback system has been employed with gated 3D thoracic MRI acquisitions. The QBH biofeedback system utilized (1) the external marker position on the abdomen using an RPM system (Real‐time Position Management, Varian) to audio‐visually guide a human subject for 2s breath‐hold at 90% exhalation position in each respiratory cycle and (2) the gated T2 weighted SPACE MR pulse sequence of 3 T Siemens MRI for 3D thoracic imaging. The improvement in the upper liver breath‐hold motion reproducibility within the gating window using the QBH biofeedback system has been assessed for a group of volunteers. Each subject underwent two imaging sessions for the assessments of the residual respiratory motion management within the gating window and respiratory motion artifacts in 3D thoracic MRI both with/without QBH biofeedback. Results: The residual upper liver motion within the gating window during MR acquisitions (∼6 minutes) has been considerably reduced using QBH biofeedback, resulting in the reduction of respiratory motion artifacts in lungs and liver of 3D thoracic MR images. Additionally, average RMSE (root mean square error) of abdomen displacement obtained from the RPM has been reduced from 2.3mm of free breathing to 0.8mm of QBH biofeedback breathing: 65% of average displacement error reduction with QBH biofeedback. Conclusion: The study demonstrated the improvement of the upper liver breath‐hold motion reproducibility using QBH biofeedback during 3D thoracic MR imaging. This system can provide clinically applicable motion management of the internal anatomy for gated radiotherapy.